[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AAX9YR1QF598RY3AEVTMAA":3,"skill-same:01M7AAX9YR1QF598RY3AEVTMAA":148},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":44,"commitSha":45,"skillDirectory":46,"forksCount":47,"repositoryHtmlUrl":48,"repositoryHomepageUrl":6,"ownerHtmlUrl":49,"ownerBlogUrl":6,"ownerDescription":50,"updatedAt":51,"install":52,"resources":65},"01M7AAX9YR1QF598RY3AEVTMAA","rds-connectivity-troubleshooter",null,"Diagnoses RDS and Aurora connection failures through a bidirectional- SG-first decision tree covering security group misconfiguration (wrong SG attached, missing inbound on the RDS SG, missing egress on the client SG), subnet group errors, DNS resolution (custom vs cluster vs writer vs reader endpoint), IAM database auth failures (token expiry, policy scope), SSL\u002FTLS issues (force_ssl, CA bundle rds-ca-2019\u002Fe2022, TLS version mismatch), max_connections reached (parameter group override, Aurora derived limit), storage-full status (auto-scaling not enabled), CPU\u002Fmemory exhaustion, read replica lag, Aurora global database replication issues, RDS Proxy connectivity, parameter group and option group conflicts. Walks describe-db-instances + describe-security-groups to a verified root cause. Emits ROOT_CAUSE_IDENTIFIED or INSUFFICIENT_DATA. Use when an application cannot reach an RDS or Aurora database, intermittent connectivity, auth\u002FTLS errors, pool exhaustion, or DNS lookup failures against an RDS endpoint.","jhkchan\u002Faws-agent-skills","jhkchan","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F1083902?v=4","01M7A2SVQYP9GQTPTDXP5JKYQ4","Jacky Chan","01M7A2SVS117YWCTZABB75HZR1",4,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: rds-connectivity-troubleshooter\ndescription: Diagnoses RDS and Aurora connection failures through a bidirectional- SG-first decision tree covering security group misconfiguration (wrong SG attached, missing inbound on the RDS SG, missing egress on the client SG), subnet group errors, DNS resolution (custom vs cluster vs writer vs reader endpoint), IAM database auth failures (token expiry, policy scope), SSL\u002FTLS issues (force_ssl, CA bundle rds-ca-2019\u002Fe2022, TLS version mismatch), max_connections reached (parameter group override, Aurora derived limit), storage-full status (auto-scaling not enabled), CPU\u002Fmemory exhaustion, read replica lag, Aurora global database replication issues, RDS Proxy connectivity, parameter group and option group conflicts. Walks describe-db-instances + describe-security-groups to a verified root cause. Emits ROOT_CAUSE_IDENTIFIED or INSUFFICIENT_DATA. Use when an application cannot reach an RDS or Aurora database, intermittent connectivity, auth\u002FTLS errors, pool exhaustion, or DNS lookup failures against an RDS endpoint.\nlicense: Apache-2.0\ncompatibility: Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). Offline symptom classification works from pasted error messages and instance metadata. Live-account diagnosis uses aws rds describe-db-instances, describe-db-clusters, describe-events, describe-db-subnet-groups, describe-db-parameter-groups, describe-option-groups, aws ec2 describe-security-groups, describe-network-acls, describe-route-tables, describe-vpc-endpoints, aws logs get-log-events \u002F filter-log-events...\nmetadata:\n  domain: aws-cloudops\n  complexity: high\n  requires_llm: 'true'\n  phase: '2'\n  supports_pipeline: 'true'\n  entry_point: 'false'\n  family: Databases\n  task_type: troubleshoot\n  skill_class: capability\n  lifecycle_status: active\n  verdict_shape: ROOT_CAUSE_IDENTIFIED | INSUFFICIENT_DATA\n  when_to_use: Diagnosing why an application cannot reach an RDS or Aurora database — connection timeout (SG, subnet, NACL, route table), authentication failure (master credentials, IAM database auth, expired password), SSL\u002FTLS handshake error (force_ssl, CA bundle, TLS version), connection refused (instance unavailable, Multi-AZ failover, maintenance window, storage-full), too many connections (max_connections, Aurora derived limit, RDS Proxy), high latency (CPU\u002Fmemory exhaustion, read replica lag), DNS resolution failure (custom endpoint, cluster endpoint, cross-region), or RDS Proxy connectivity issues.\n  when_not_to_use: Query-level performance tuning (use Performance Insights + slow query log); RDS instance configuration posture audits (use rds-instance-auditor); Aurora failover automation (use aurora-failover-operator); RDS backup and restore operations (use rds-backup-restore-operator); parameter group authoring (use rds-parameter-group-deployer).\n  activation_triggers: RDS connection timeout, cannot connect to RDS, Aurora connection failed, authentication failed for database, FATAL password authentication failed, Access denied for user, SSL connection required, TLS handshake failed RDS, connection refused RDS, too many connections RDS, FATAL too many connections, RDS storage-full, RDS CPU exhaustion, read replica lag, Aurora global database replication, RDS Proxy unreachable, parameter group max_connections override, option group conflict, RDS DNS not resolving\n  invocation_schema: 'Input: either (a) a symptom description (the error string or observed behaviour, the DB instance or cluster identifier), OR (b) a live-account scenario where the agent runs aws rds describe-db-instances \u002F describe-db-clusters \u002F describe-events \u002F describe-security-groups to gather evidence. Output: a deterministic TARGET \u002F VERDICT \u002F ROOT_CAUSE \u002F REASON \u002F EVIDENCE \u002F REMEDIATION block where VERDICT ∈ {ROOT_CAUSE_IDENTIFIED, INSUFFICIENT_DATA} and ROOT_CAUSE names the specific failure category (NETWORK_SG_INBOUND \u002F NETWORK_SG_EGRESS \u002F NETWORK_NACL \u002F NETWORK_ROUTE \u002F NETWORK_SUBNET_GROUP \u002F NETWORK_CROSS_VPC \u002F DNS_CUSTOM_ENDPOINT \u002F DNS_CLUSTER_ENDPOINT \u002F AUTH_CREDENTIALS \u002F AUTH_IAM_DB \u002F AUTH_EXPIRED \u002F TLS_CONFIG \u002F TLS_CERT \u002F INSTANCE_UNAVAILABLE \u002F INSTANCE_FAILOVER \u002F INSTANCE_MAINTENANCE \u002F INSTANCE_STORAGE_FULL \u002F CAPACITY_MAX_CONNECTIONS \u002F CAPACITY_DERIVED_LIMIT \u002F CAPACITY_PROXY \u002F RESOURCE_CPU \u002F RESOURCE_MEMORY \u002F REPLICATION_LAG \u002F GLOBAL_DB_REPLICATION \u002F PARAM_GROUP_OVERRIDE \u002F OPTION_GROUP_CONFLICT \u002F UNKNOWN).'\n  version: 0.1.0\n  author: Jacky Chan — AWS Community Builder\n  keywords: RDS, Aurora, connection timeout, security group, bidirectional SG, subnet group, DB subnet group, DNS resolution, custom endpoint, cluster endpoint, writer endpoint, reader endpoint, IAM database auth, SSL, TLS, force_ssl, require_secure_transport, CA bundle, rds-ca-2019, rds-ca-e2022, max_connections, Aurora derived limit, storage-full, storage auto-scaling, CPU exhaustion, memory exhaustion, read replica lag, Aurora global database, RDS Proxy, parameter group, option group\n  tags: rds, aurora, databases, troubleshoot, connectivity, security-group, tls, iam-db-auth, storage-full\n---\n\n# RDS Connectivity Troubleshooter\n\n## Quick navigation\n\n| Symptom \u002F signal | Jump | Likely ROOT_CAUSE |\n|---|---|---|\n| `connection timed out` \u002F `Could not connect to server` | Step 1 | `NETWORK_SG_INBOUND` (most common) or `NETWORK_SG_EGRESS` |\n| `connection refused` (RST returned) | Step 2 | `INSTANCE_UNAVAILABLE` \u002F `INSTANCE_FAILOVER` \u002F `INSTANCE_MAINTENANCE` \u002F `INSTANCE_STORAGE_FULL` |\n| `FATAL: password authentication failed` \u002F `Access denied for user` | Step 3 | `AUTH_CREDENTIALS` or `AUTH_IAM_DB` |\n| `SSL connection required` \u002F `TLS handshake failed` | Step 4 | `TLS_CONFIG` (`force_ssl` \u002F `require_secure_transport`) or `TLS_CERT` |\n| `FATAL: too many connections` \u002F pool exhaustion | Step 5 | `CAPACITY_MAX_CONNECTIONS` \u002F `CAPACITY_DERIVED_LIMIT` \u002F `CAPACITY_PROXY` |\n| `could not translate host name` \u002F `nodename nor servname provided` | Step 6 | `DNS_CUSTOM_ENDPOINT` \u002F `DNS_CLUSTER_ENDPOINT` |\n| writer endpoint routes to a reader after failover | Step 6 | `DNS_CLUSTER_ENDPOINT` (Aurora writer DNS update lag) |\n| `storage-full` status in describe-db-instances | Step 2 | `INSTANCE_STORAGE_FULL` (auto-scaling not enabled) |\n| Aurora global database secondary lag > 1 minute | Step 7 | `GLOBAL_DB_REPLICATION` |\n| read replica `ReplicaLag` > 30s and rising | Step 7 | `REPLICATION_LAG` |\n| RDS Proxy `could not connect to proxy` | Step 8 | `CAPACITY_PROXY` (proxy SG \u002F secrets) |\n| parameter group change preceded the failure | Step 9 | `PARAM_GROUP_OVERRIDE` |\n| option group change preceded the failure | Step 9 | `OPTION_GROUP_CONFLICT` |\n| Symptom does not match any row, evidence incomplete | Step 10 | `INSUFFICIENT_DATA` |\n\n## STRICT output contract\n\nEvery invocation MUST emit exactly one diagnostic block as the final\nanswer. The block is machine-parsable; any drift breaks the eval\nharness and the downstream orchestrator.\n\n```text\nTARGET: \u003Cdb-instance-identifier \u002F cluster-identifier \u002F endpoint>\nVERDICT: ROOT_CAUSE_IDENTIFIED | INSUFFICIENT_DATA\nROOT_CAUSE: \u003CCATEGORY_NAME>\nREASON: \u003C1-3 sentences naming the failing config element + the probe\n  that proves it>\nEVIDENCE:\n  - \u003Cobserved signal — error string \u002F instance status \u002F metric>\n  - \u003Cfailing probe — CLI command and the specific output line that\n    confirms the cause>\n  - \u003Cpassing probes — categories ruled out with one-line justification>\nREMEDIATION:\n  1. \u003Cspecific action with CLI command>\n  2. \u003Cverification command after the fix>\nCONFIRM: Before any state-changing CLI, emit and await operator\n  approval: \"CONFIRM: About to \u003Caction> on \u003Cresource>. Proceed? (yes\u002Fno)\"\n```\n\nRules:\n\n- `VERDICT: ROOT_CAUSE_IDENTIFIED` requires at least one FAILING probe\n  that matches the symptom — never a process-of-elimination answer.\n- `VERDICT: INSUFFICIENT_DATA` MUST list the missing inputs and the\n  next probe to run; it is never a dead-end.\n- `ROOT_CAUSE` is a single UPPER_SNAKE_CASE category name from the set\n  above — free-text root causes are forbidden.\n- `EVIDENCE` lists BOTH the failing probe AND the categories ruled out.\n- One block per target; multi-target incidents separate blocks with `---`.\n\n## NEVER\n\n- **NEVER** declare `ROOT_CAUSE_IDENTIFIED` without a failing probe\n  that matches the symptom. A `connection timed out` alone is a\n  symptom; the failing probe is the specific SG rule, NACL rule, or\n  route that drops the packet.\n- **NEVER** check only the RDS security group for a connection timeout.\n  Security group rules are BIDIRECTIONAL — the client's SG must allow\n  egress on the DB port AND the RDS SG must allow ingress from the\n  client's SG (or CIDR). Operators who fix the RDS-side rule and skip\n  the client-side egress rule see the timeout persist and conclude\n  \"security groups are fine.\" Always inspect BOTH SGs.\n- **NEVER** assume the Aurora cluster endpoint points at the writer.\n  The cluster endpoint follows the writer after a failover, but DNS\n  caches (resolver, JVM, application) can serve a stale record for the\n  TTL window (default 1s for the cluster endpoint, longer for cached\n  resolvers). After a failover, the old writer becomes a reader and\n  the cluster endpoint updates — but a write to a reader throws\n  `ERROR: cannot execute INSERT in a read-only transaction`. Always\n  confirm what the endpoint currently resolves to.\n- **NEVER** treat `storage-full` as a capacity metric issue. When an\n  RDS instance hits `storage-full`, it STOPS accepting writes — the\n  application sees connection errors, query timeouts, or `disk full`\n  errors. The instance status in `describe-db-instances` shows\n  `StorageStatus: storage-full`. The fix is enabling storage\n  auto-scaling (or allocating more storage); raising\n  `max_connections` does nothing.\n- **NEVER** confuse IAM database auth with the master credentials.\n  IAM database auth generates a short-lived (15-minute) SigV4 token\n  used as the password; the master credentials are static. A\n  `password authentication failed` for an IAM-auth flow means the\n  token expired OR the IAM policy lacks `rds-db:connect` OR the\n  database user was not created with `AWSAuthenticationPlugin`. Adding\n  `rds-db:connect` to a user connecting with the master password is a\n  no-op.\n- **NEVER** recommend disabling `force_ssl` \u002F `require_secure_transport`\n  to fix a TLS error in production. The TLS error is a SYMPTOM of a\n  legitimate mismatch (wrong CA bundle, TLS 1.0 against a TLS 1.2+\n  instance, missing `sslmode` in the connection string). Disabling\n  encryption is a security regression.\n- **NEVER** assume a read replica is queryable. A replica in\n  `Replicating` state but with `ReplicaLag` climbing is falling behind\n  — reads return stale data and CPU is saturated by replication\n  replay. Always check `ReplicaLag` before routing reads to a replica.\n\n## Expert heuristic\n\nThe three expert-heuristic deep dives (bidirectional SG check order, Aurora\nwriter\u002Freader routing, storage-full) moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## Configuration dependency graph\n\nThe connectivity configuration dependency graph (ASCII) and its\nread-top-down guidance moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## Mindset\n\nAn RDS connection failure is almost never a database-engine problem.\nThe engine is fine; the broken thing is a security group rule, a\nsubnet group misconfiguration, an expired IAM token, a TLS mismatch, a\nstorage-full instance, or a stale DNS cache. Senior database engineers\nstart with `describe-db-instances` (for the instance status, endpoint,\nSG list, and subnet group) and `describe-security-groups` (for both\nthe RDS SG and the client SG), then drill into engine logs only after\nthe network and auth layers are confirmed clean. Operators who start\nwith the slow query log miss a `storage-full` status that fired\nBEFORE the query reached the engine.\n\n## Process — bidirectional-SG-first diagnostic tree\n\n### Step 0: Pre-flight — gather instance and client state\n\nInstance\u002Fcluster\u002Fevents\u002Fclient-state gathering commands (probes 1-4) moved verbatim to\n[references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\n\nShort-circuit cases that mimic connectivity failure:\n\n| Signal | Effect on diagnosis |\n|---|---|\n| `DBInstanceStatus: deleted` \u002F `deleting` | Instance is gone; emit `INSUFFICIENT_DATA` listing the deleted instance. |\n| `DBInstanceStatus: modifying` | Instance is in a maintenance window; connections may drop briefly. Wait and re-check. |\n| `DBInstanceStatus: incompatible-restore` \u002F `incompatible-parameters` | Parameter group change broke the engine; fold to Step 9. |\n| `StorageStatus: storage-full` | Instance stopped accepting writes; fold to Step 2 (`INSTANCE_STORAGE_FULL`). |\n| AWS Health event `AWS_RDS_SERVICE` open in the region | Fold to AWS-side; surface the event ARN. Do NOT continue customer-side diagnosis. |\n\nIf the input lacks the instance identifier or the symptom, emit\n`INSUFFICIENT_DATA` listing the missing fields.\n\n### Step 1: Connection timeout — bidirectional security group\n\nSymptom: `connection timed out`, `Could not connect to server`,\n`TimeoutExpired`, or the TCP SYN goes unanswered. This is a packet\ndrop, NOT an auth or TLS issue.\n\n#### 1a: RDS SG inbound (most common)\n\n```bash\naws ec2 describe-security-groups --group-ids \u003Crds-sg-id> --output json | \\\n  jq('.SecurityGroups[0].IpPermissions')\n```\n\nThe RDS SG must have an inbound rule allowing the client's SG (or\nCIDR) on the DB port (5432 for Postgres, 3306 for MySQL, 1433 for\nSQL Server, 1521 for Oracle, 5439 for Redshift). If the rule is\nmissing or scoped to a different SG \u002F CIDR, **ROOT_CAUSE_IDENTIFIED**\nwith `ROOT_CAUSE: NETWORK_SG_INBOUND`.\n\n#### 1b: Client SG egress (the other half of the bidirectional check)\n\n```bash\naws ec2 describe-security-groups --group-ids \u003Cclient-sg-id> --output json | \\\n  jq('.SecurityGroups[0].IpPermissionsEgress')\n```\n\nThe client SG must allow egress to the RDS endpoint on the DB port.\nThe default egress is allow-all (0.0.0.0\u002F0); a locked-down SG with a\nrestricted egress (e.g., only 443) silently drops the DB-bound SYN. If\nthe egress rule is missing, **ROOT_CAUSE_IDENTIFIED** with\n`ROOT_CAUSE: NETWORK_SG_EGRESS`.\n\n#### 1c: NACL and route table (deeper network)\n\nIf both SGs allow the flow, check the NACL on both subnets (NACLs are\nstateless — both inbound and outbound rules must allow the flow in\nboth directions) and the route table (for cross-VPC peering \u002F TGW\nroutes). Fold to `NETWORK_NACL` or `NETWORK_ROUTE`.\n\n### Step 2: Connection refused — instance-level failure\n\nSymptom: `connection refused` (RST returned, not a timeout). The\nnetwork path is open but the DB engine is not accepting connections.\n\n| describe-db-instances signal | ROOT_CAUSE |\n|---|---|\n| `DBInstanceStatus: available` but `StorageStatus: storage-full` | `INSTANCE_STORAGE_FULL` — enable storage auto-scaling |\n| `DBInstanceStatus: modifying` | `INSTANCE_MAINTENANCE` — wait for the window to complete |\n| Recent event `DB instance restarted` \u002F `Multi-AZ failover completed` | `INSTANCE_FAILOVER` — the writer changed; application DNS cache may be stale |\n| `DBInstanceStatus: stopped` \u002F `starting` | `INSTANCE_UNAVAILABLE` — start the instance |\n| `DBInstanceStatus: incompatible-*` | Fold to Step 9 (parameter or option group) |\n\nFor `storage-full`: confirm with `describe-db-instances` showing\n`StorageStatus: storage-full`. The fix is:\n\n```bash\naws rds modify-db-instance --db-instance-identifier \u003Cid> \\\n  --storage-auto-scaling --max-allocated-storage \u003CN> --apply-immediately\n```\n\n### Step 3: Authentication failure — credentials vs IAM DB auth\n\nSymptom: `FATAL: password authentication failed for user \"...\"`,\n`Access denied for user`, or `28000: authentication failed`.\n\n#### 3a: Master \u002F static credentials\n\nIf the connection uses a static password (Secrets Manager, hardcoded,\nor env var), the credentials are wrong, rotated, or stale. Check\nSecrets Manager:\n\n```bash\naws secretsmanager get-secret-value --secret-id \u003Csecret-id> \\\n  --output json | jq('.SecretString')\n```\n\nCompare the username \u002F password against what the application is using.\nIf the secret was rotated and the application cached the old value,\n**ROOT_CAUSE_IDENTIFIED** with `ROOT_CAUSE: AUTH_CREDENTIALS`.\n\n#### 3b: IAM database auth\n\nIf `IAMDatabaseAuthenticationEnabled: true` and the application uses\nan SigV4 token as the password, the token expires after 15 minutes.\nCheck:\n\n1. The IAM policy attached to the application's role must include\n   `rds-db:connect` on the DB instance ARN.\n2. The database user must exist with\n   `AWSAuthenticationPlugin: rds_iam` (Postgres) or\n   `AWSAuthenticationPlugin: mysql_native_password` (MySQL).\n3. The token must be regenerated before each connection (or at least\n   every 14 minutes).\n\n```bash\n# Verify the IAM policy grants rds-db:connect\naws iam simulate-principal-policy \\\n  --policy-source-arn \u003Capp-role-arn> \\\n  --action-names rds-db:connect \\\n  --resource-arns arn:aws:rds-db:\u003Cregion>:\u003Caccount>:dbuser:\u003Cinstance>\u002F\u003Cuser> \\\n  --output json\n```\n\nIf `implicitDeny`, **ROOT_CAUSE_IDENTIFIED** with\n`ROOT_CAUSE: AUTH_IAM_DB`. If the token is older than 15 minutes,\n`ROOT_CAUSE: AUTH_EXPIRED`.\n\n### Step 4: SSL \u002F TLS errors\n\nSymptom: `SSL connection required`, `TLS handshake failed`,\n`no SSL capabilities`, or `certificate verify failed`.\n\n#### 4a: force_ssl \u002F require_secure_transport\n\n```bash\naws rds describe-db-instances --db-instance-identifier \u003Cid> \\\n  --output json | jq('.DBInstances[0] | .DBParameterGroups')\n```\n\nIf the parameter group has `rds.force_ssl: 1` (Postgres) or\n`require_secure_transport: 1` (MySQL) and the client connects without\n`sslmode=require` (or the MySQL equivalent), the connection is\nrejected. **ROOT_CAUSE_IDENTIFIED** with `ROOT_CAUSE: TLS_CONFIG`.\n\n#### 4b: CA bundle mismatch\n\n```bash\naws rds describe-db-instances --db-instance-identifier \u003Cid> \\\n  --output json | jq('.DBInstances[0].CACertificateIdentifier')\n```\n\nIf the instance uses `rds-ca-e2022` (the recommended rotation) but the\nclient's trust store has `rds-ca-2019`, the TLS handshake fails with\n`certificate verify failed`. **ROOT_CAUSE_IDENTIFIED** with\n`ROOT_CAUSE: TLS_CERT`. Fix: update the client trust store with the\nnew CA bundle from the AWS RDS docs.\n\n### Step 5: Too many connections — capacity\n\nSymptom: `FATAL: too many connections`,\n`FATAL: remaining connection slots are reserved for non-replication\nsuperuser connections`, or connection pool exhaustion at the client.\n\n#### 5a: max_connections (RDS for MySQL \u002F Postgres)\n\n```bash\naws rds describe-db-instances --db-instance-identifier \u003Cid> \\\n  --output json | jq('.DBInstances[0].DBParameterGroups')\n```\n\nIf a parameter group overrides `max_connections` below the default,\n**ROOT_CAUSE_IDENTIFIED** with\n`ROOT_CAUSE: CAPACITY_MAX_CONNECTIONS` (or\n`PARAM_GROUP_OVERRIDE` if the override is the root issue).\n\n#### 5b: Aurora derived limit\n\nAurora derives `max_connections` from the instance class\n(`LEAST({DBInstanceClassMemory\u002F9531392, 5000})` for Aurora\nPostgreSQL). Upsizing the instance class raises the limit\nautomatically. If the application exceeds the derived limit,\n**ROOT_CAUSE_IDENTIFIED** with `ROOT_CAUSE: CAPACITY_DERIVED_LIMIT`.\nFix: upsize the instance class OR add an RDS Proxy to pool\nconnections.\n\n#### 5c: RDS Proxy not absorbing the load\n\nIf an RDS Proxy is configured but the application still hits\n`too many connections`, the proxy may be misconfigured (wrong target\nSG, wrong secrets, or the application is bypassing the proxy\nendpoint). Fold to Step 8.\n\n### Step 6: DNS resolution — custom vs cluster endpoint\n\nSymptom: `could not translate host name`, `nodename nor servname`,\nor the application connects to the wrong instance after a failover.\n\n#### 6a: Custom endpoint vs cluster endpoint\n\nAurora custom endpoints route to a subset of instances; the cluster\nendpoint follows the writer; the reader endpoint round-robins across\nreaders. If the application uses a custom endpoint for writes but the\ncustom endpoint only includes readers, every write fails with\n`cannot execute INSERT in a read-only transaction`.\n\n```bash\naws rds describe-db-clusters --db-cluster-identifier \u003Ccluster> \\\n  --output json | jq('.DBClusters[0] | {\n    CustomEndpoints: .CustomEndpoints,\n    Endpoint, ReaderEndpoint,\n    Members: [.DBClusterMembers[] | {DBInstanceIdentifier,\n      IsClusterWriter}]}')\n```\n\nConfirm the endpoint includes the writer. **ROOT_CAUSE_IDENTIFIED**\nwith `ROOT_CAUSE: DNS_CUSTOM_ENDPOINT` or `DNS_CLUSTER_ENDPOINT`.\n\n#### 6b: Stale DNS after failover\n\nStale-DNS-after-failover guidance moved verbatim to\n[references\u002Fconnectivity-layer-reference.md](references\u002Fconnectivity-layer-reference.md).\n\n### Step 7: Replication lag — read replica and global database\n\n#### 7a: Read replica lag\n\n```bash\naws cloudwatch get-metric-statistics --namespace AWS\u002FRDS \\\n  --metric-name ReplicaLag \\\n  --dimensions Name=DBInstanceIdentifier,Value=\u003Creplica-id> \\\n  --start-time $(date -d '-1 hour' +%FT%TZ) --end-time $(date +%FT%TZ) \\\n  --period 300 --statistics Average,Maximum --output json\n```\n\n`ReplicaLag` > 30s and rising indicates the replica cannot keep up\nwith the source. Reads return stale data. **ROOT_CAUSE_IDENTIFIED**\nwith `ROOT_CAUSE: REPLICATION_LAG`. Fix: upsize the replica, reduce\nwrite load on the source, or route reads elsewhere.\n\n#### 7b: Aurora global database replication\n\nFor Aurora global databases, the secondary cluster replicates from\nthe primary over a dedicated channel. If the channel is interrupted\n(network, AWS-side issue) or the secondary cannot keep up,\n`ReplicationLag` on the secondary climbs. Check the secondary cluster\nstatus and the AWS Health dashboard for cross-region issues.\n**ROOT_CAUSE_IDENTIFIED** with `ROOT_CAUSE: GLOBAL_DB_REPLICATION`.\n\n### Step 8: RDS Proxy connectivity\n\nRDS Proxy probe commands and the common proxy failure pattern table moved verbatim to\n[references\u002Fengine-auth-reference.md](references\u002Fengine-auth-reference.md).\n\n### Step 9: Parameter group and option group misconfiguration\n\n#### 9a: Parameter group override\n\nThe parameter-override culprit table moved verbatim to\n[references\u002Fengine-auth-reference.md](references\u002Fengine-auth-reference.md).\n\n#### 9b: Option group conflict\n\nOption-group conflict patterns moved verbatim to\n[references\u002Fengine-auth-reference.md](references\u002Fengine-auth-reference.md).\n\n### Step 10: INSUFFICIENT_DATA — when to bail\n\nEmit `INSUFFICIENT_DATA` when:\n\n- The input lacks the instance identifier or endpoint and the symptom\n  is too generic.\n- A probe requires operator input (e.g., the application's connection\n  string is unknown).\n- The symptom matches no row in the quick navigation table and the\n  describe-db-instances output is empty or stale.\n- An AWS Health event is OPEN and may be the cause.\n\nALWAYS list the missing inputs and the exact next probe to run.\n\n## Output format\n\nSee the STRICT output contract above. Two worked examples follow.\n\n### Worked example — connection timeout, RDS SG missing inbound\n\n```text\nTARGET: db-prod-postgres (endpoint: db-prod-postgres.abc123.us-east-1.rds.amazonaws.com:5432)\nVERDICT: ROOT_CAUSE_IDENTIFIED\nROOT_CAUSE: NETWORK_SG_INBOUND\nREASON: The RDS SG (sg-rds-aaa) has no inbound rule allowing the\n  application's SG (sg-app-bbb) on port 5432. The only inbound rule\n  permits sg-old-ccc (a decommissioned SG). The client SG egress is\n  allow-all (0.0.0.0\u002F0), so the packet reaches the RDS subnet but is\n  dropped at the RDS SG.\nEVIDENCE:\n  - Probe: aws ec2 describe-security-groups --group-ids sg-rds-aaa\n    returns IpPermissions with one rule: port 5432 from sg-old-ccc —\n    sg-app-bbb is absent.\n  - Probe: aws ec2 describe-security-groups --group-ids sg-app-bbb\n    returns IpPermissionsEgress: 0.0.0.0\u002F0 all ports (egress is open).\n  - Passing: describe-db-instances shows DBInstanceStatus: available,\n    StorageStatus: healthy (instance is not the issue); the client is\n    in the same VPC (route table is not the issue).\nREMEDIATION:\n  1. Add an inbound rule to sg-rds-aaa allowing sg-app-bbb on 5432:\n     aws ec2 authorize-security-group-ingress --group-id sg-rds-aaa \\\n       --protocol tcp --port 5432 --source-security-group-id sg-app-bbb\n  2. Verify from the client host: psql -h \u003Cendpoint> -U \u003Cuser> -d \u003Cdb>\n     (should connect within seconds).\nCONFIRM: Before adding the rule, emit and await:\n  \"CONFIRM: About to add inbound 5432 from sg-app-bbb to sg-rds-aaa.\n   Proceed? (yes\u002Fno)\"\n```\n\n### Worked example — INSUFFICIENT_DATA\n\nThe INSUFFICIENT_DATA worked example moved verbatim to\n[references\u002Fworked-examples.md](references\u002Fworked-examples.md).\n\n## Pre-flight safety checks\n\nPre-flight safety checks (confirm gate, read-only-first, disruption notes,\nbatch limit) moved verbatim to [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\n\n## Remediation guidance\n\n| ROOT_CAUSE | Specific fix |\n|---|---|\n| `NETWORK_SG_INBOUND` | Add inbound rule to the RDS SG for the client SG \u002F CIDR on the DB port. |\n| `NETWORK_SG_EGRESS` | Add egress rule to the client SG for the RDS endpoint on the DB port. |\n| `NETWORK_NACL` \u002F `NETWORK_ROUTE` | Add NACL \u002F route table rule for bidirectional flow. |\n| `NETWORK_SUBNET_GROUP` | Add a subnet in the client's AZ to the DB subnet group. |\n| `INSTANCE_STORAGE_FULL` | Enable storage auto-scaling: `modify-db-instance --storage-auto-scaling --max-allocated-storage \u003CN>`. |\n| `INSTANCE_FAILOVER` | Wait for the cluster endpoint DNS to update; flush the app DNS cache. |\n| `AUTH_CREDENTIALS` | Rotate \u002F update the secret in Secrets Manager; refresh the app's cached value. |\n| `AUTH_IAM_DB` | Add `rds-db:connect` to the app role; create the DB user with `AWSAuthenticationPlugin`. |\n| `TLS_CONFIG` | Add `sslmode=require` (Postgres) or `requireSSL=true` (MySQL) to the connection string. |\n| `TLS_CERT` | Update the client trust store with the CA bundle matching `CACertificateIdentifier`. |\n| `CAPACITY_MAX_CONNECTIONS` \u002F `CAPACITY_DERIVED_LIMIT` | Upsize the instance class OR add an RDS Proxy to pool connections. |\n| `CAPACITY_PROXY` | Fix the proxy SG, secrets, or application connection string. |\n| `REPLICATION_LAG` | Upsize the replica, reduce source write load, or route reads elsewhere. |\n| `GLOBAL_DB_REPLICATION` | Check the secondary cluster status and AWS Health; recover or re-establish the channel. |\n| `DNS_CUSTOM_ENDPOINT` \u002F `DNS_CLUSTER_ENDPOINT` | Add the writer to the custom endpoint; flush DNS; use the writer endpoint for writes. |\n| `PARAM_GROUP_OVERRIDE` | Revert or correct the parameter; reboot if required. |\n| `OPTION_GROUP_CONFLICT` | Resolve the option conflict; re-apply the option group. |\n\n## References (load on demand)\n\n- [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md) — expert-heuristic deep dives (bidirectional SG check order, Aurora writer\u002Freader routing, storage-full) and the configuration dependency graph (moved from this file)\n- [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md) — Step 0 pre-flight gathering commands (probes 1-4) and the pre-flight safety checks (moved from this file)\n- [references\u002Fworked-examples.md](references\u002Fworked-examples.md) — the INSUFFICIENT_DATA worked example; the primary connection-timeout example stays in this file\n- [references\u002Fengine-auth-reference.md](references\u002Fengine-auth-reference.md) — IAM DB auth, TLS\u002FCA rotation, parameter and option group, storage-full, proxy, DNS detail; extended with Step 8 proxy probes, the 9a parameter-override table, and 9b option-group conflicts\n- [references\u002Fconnectivity-layer-reference.md](references\u002Fconnectivity-layer-reference.md) — engine port registry, SG\u002FNACL evaluation rules, Aurora endpoint behaviour, cross-VPC matrix, AWS Health event categories; extended with the 6b stale-DNS-after-failover guidance\n\n## Domain\n\nAWS CloudOps \u002F RDS & Aurora Databases, Connectivity Diagnostics, VPC\nNetworking for Databases, IAM Database Authentication, TLS \u002F SSL,\nStorage Auto-Scaling, and Read Replica \u002F Global Database Replication.\n\n## AWS documentation\n\n- RDS troubleshooting — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FCHAP_Troubleshooting.html\n- Security group rules — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_VPC.WorkingWithRDSInstanceinaVPC.html\n- IAM database auth — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUsingWithRDS.IAMDBAuth.html\n- SSL\u002FTLS — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUsingWithRDS.SSL.html\n- Storage auto-scaling — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_PIOPS.StorageTypes.html#USER_PIOPS.Autoscaling\n- Aurora endpoints — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002FAurora.Overview.Endpoints.html\n- Aurora global databases — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002Faurora-global-database.html\n- RDS Proxy — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002Frds-proxy.html\n- Aurora failover — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002FConcepts.AuroraHighAvailability.html\n","# RDS Connectivity Troubleshooter\n\n## Quick navigation\n\n| Symptom \u002F signal | Jump | Likely ROOT_CAUSE |\n|---|---|---|\n| `connection timed out` \u002F `Could not connect to server` | Step 1 | `NETWORK_SG_INBOUND` (most common) or `NETWORK_SG_EGRESS` |\n| `connection refused` (RST returned) | Step 2 | `INSTANCE_UNAVAILABLE` \u002F `INSTANCE_FAILOVER` \u002F `INSTANCE_MAINTENANCE` \u002F `INSTANCE_STORAGE_FULL` |\n| `FATAL: password authentication failed` \u002F `Access denied for user` | Step 3 | `AUTH_CREDENTIALS` or `AUTH_IAM_DB` |\n| `SSL connection required` \u002F `TLS handshake failed` | Step 4 | `TLS_CONFIG` (`force_ssl` \u002F `require_secure_transport`) or `TLS_CERT` |\n| `FATAL: too many connections` \u002F pool exhaustion | Step 5 | `CAPACITY_MAX_CONNECTIONS` \u002F `CAPACITY_DERIVED_LIMIT` \u002F `CAPACITY_PROXY` |\n| `could not translate host name` \u002F `nodename nor servname provided` | Step 6 | `DNS_CUSTOM_ENDPOINT` \u002F `DNS_CLUSTER_ENDPOINT` |\n| writer endpoint routes to a reader after failover | Step 6 | `DNS_CLUSTER_ENDPOINT` (Aurora writer DNS update lag) |\n| `storage-full` status in describe-db-instances | Step 2 | `INSTANCE_STORAGE_FULL` (auto-scaling not enabled) |\n| Aurora global database secondary lag > 1 minute | Step 7 | `GLOBAL_DB_REPLICATION` |\n| read replica `ReplicaLag` > 30s and rising | Step 7 | `REPLICATION_LAG` |\n| RDS Proxy `could not connect to proxy` | Step 8 | `CAPACITY_PROXY` (proxy SG \u002F secrets) |\n| parameter group change preceded the failure | Step 9 | `PARAM_GROUP_OVERRIDE` |\n| option group change preceded the failure | Step 9 | `OPTION_GROUP_CONFLICT` |\n| Symptom does not match any row, evidence incomplete | Step 10 | `INSUFFICIENT_DATA` |\n\n## STRICT output contract\n\nEvery invocation MUST emit exactly one diagnostic block as the final\nanswer. The block is machine-parsable; any drift breaks the eval\nharness and the downstream orchestrator.\n\n```text\nTARGET: \u003Cdb-instance-identifier \u002F cluster-identifier \u002F endpoint>\nVERDICT: ROOT_CAUSE_IDENTIFIED | INSUFFICIENT_DATA\nROOT_CAUSE: \u003CCATEGORY_NAME>\nREASON: \u003C1-3 sentences naming the failing config element + the probe\n  that proves it>\nEVIDENCE:\n  - \u003Cobserved signal — error string \u002F instance status \u002F metric>\n  - \u003Cfailing probe — CLI command and the specific output line that\n    confirms the cause>\n  - \u003Cpassing probes — categories ruled out with one-line justification>\nREMEDIATION:\n  1. \u003Cspecific action with CLI command>\n  2. \u003Cverification command after the fix>\nCONFIRM: Before any state-changing CLI, emit and await operator\n  approval: \"CONFIRM: About to \u003Caction> on \u003Cresource>. Proceed? (yes\u002Fno)\"\n```\n\nRules:\n\n- `VERDICT: ROOT_CAUSE_IDENTIFIED` requires at least one FAILING probe\n  that matches the symptom — never a process-of-elimination answer.\n- `VERDICT: INSUFFICIENT_DATA` MUST list the missing inputs and the\n  next probe to run; it is never a dead-end.\n- `ROOT_CAUSE` is a single UPPER_SNAKE_CASE category name from the set\n  above — free-text root causes are forbidden.\n- `EVIDENCE` lists BOTH the failing probe AND the categories ruled out.\n- One block per target; multi-target incidents separate blocks with `---`.\n\n## NEVER\n\n- **NEVER** declare `ROOT_CAUSE_IDENTIFIED` without a failing probe\n  that matches the symptom. A `connection timed out` alone is a\n  symptom; the failing probe is the specific SG rule, NACL rule, or\n  route that drops the packet.\n- **NEVER** check only the RDS security group for a connection timeout.\n  Security group rules are BIDIRECTIONAL — the client's SG must allow\n  egress on the DB port AND the RDS SG must allow ingress from the\n  client's SG (or CIDR). Operators who fix the RDS-side rule and skip\n  the client-side egress rule see the timeout persist and conclude\n  \"security groups are fine.\" Always inspect BOTH SGs.\n- **NEVER** assume the Aurora cluster endpoint points at the writer.\n  The cluster endpoint follows the writer after a failover, but DNS\n  caches (resolver, JVM, application) can serve a stale record for the\n  TTL window (default 1s for the cluster endpoint, longer for cached\n  resolvers). After a failover, the old writer becomes a reader and\n  the cluster endpoint updates — but a write to a reader throws\n  `ERROR: cannot execute INSERT in a read-only transaction`. Always\n  confirm what the endpoint currently resolves to.\n- **NEVER** treat `storage-full` as a capacity metric issue. When an\n  RDS instance hits `storage-full`, it STOPS accepting writes — the\n  application sees connection errors, query timeouts, or `disk full`\n  errors. The instance status in `describe-db-instances` shows\n  `StorageStatus: storage-full`. The fix is enabling storage\n  auto-scaling (or allocating more storage); raising\n  `max_connections` does nothing.\n- **NEVER** confuse IAM database auth with the master credentials.\n  IAM database auth generates a short-lived (15-minute) SigV4 token\n  used as the password; the master credentials are static. A\n  `password authentication failed` for an IAM-auth flow means the\n  token expired OR the IAM policy lacks `rds-db:connect` OR the\n  database user was not created with `AWSAuthenticationPlugin`. Adding\n  `rds-db:connect` to a user connecting with the master password is a\n  no-op.\n- **NEVER** recommend disabling `force_ssl` \u002F `require_secure_transport`\n  to fix a TLS error in production. The TLS error is a SYMPTOM of a\n  legitimate mismatch (wrong CA bundle, TLS 1.0 against a TLS 1.2+\n  instance, missing `sslmode` in the connection string). Disabling\n  encryption is a security regression.\n- **NEVER** assume a read replica is queryable. A replica in\n  `Replicating` state but with `ReplicaLag` climbing is falling behind\n  — reads return stale data and CPU is saturated by replication\n  replay. Always check `ReplicaLag` before routing reads to a replica.\n\n## Expert heuristic\n\nThe three expert-heuristic deep dives (bidirectional SG check order, Aurora\nwriter\u002Freader routing, storage-full) moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## Configuration dependency graph\n\nThe connectivity configuration dependency graph (ASCII) and its\nread-top-down guidance moved verbatim to [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md).\n\n## Mindset\n\nAn RDS connection failure is almost never a database-engine problem.\nThe engine is fine; the broken thing is a security group rule, a\nsubnet group misconfiguration, an expired IAM token, a TLS mismatch, a\nstorage-full instance, or a stale DNS cache. Senior database engineers\nstart with `describe-db-instances` (for the instance status, endpoint,\nSG list, and subnet group) and `describe-security-groups` (for both\nthe RDS SG and the client SG), then drill into engine logs only after\nthe network and auth layers are confirmed clean. Operators who start\nwith the slow query log miss a `storage-full` status that fired\nBEFORE the query reached the engine.\n\n## Process — bidirectional-SG-first diagnostic tree\n\n### Step 0: Pre-flight — gather instance and client state\n\nInstance\u002Fcluster\u002Fevents\u002Fclient-state gathering commands (probes 1-4) moved verbatim to\n[references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\n\nShort-circuit cases that mimic connectivity failure:\n\n| Signal | Effect on diagnosis |\n|---|---|\n| `DBInstanceStatus: deleted` \u002F `deleting` | Instance is gone; emit `INSUFFICIENT_DATA` listing the deleted instance. |\n| `DBInstanceStatus: modifying` | Instance is in a maintenance window; connections may drop briefly. Wait and re-check. |\n| `DBInstanceStatus: incompatible-restore` \u002F `incompatible-parameters` | Parameter group change broke the engine; fold to Step 9. |\n| `StorageStatus: storage-full` | Instance stopped accepting writes; fold to Step 2 (`INSTANCE_STORAGE_FULL`). |\n| AWS Health event `AWS_RDS_SERVICE` open in the region | Fold to AWS-side; surface the event ARN. Do NOT continue customer-side diagnosis. |\n\nIf the input lacks the instance identifier or the symptom, emit\n`INSUFFICIENT_DATA` listing the missing fields.\n\n### Step 1: Connection timeout — bidirectional security group\n\nSymptom: `connection timed out`, `Could not connect to server`,\n`TimeoutExpired`, or the TCP SYN goes unanswered. This is a packet\ndrop, NOT an auth or TLS issue.\n\n#### 1a: RDS SG inbound (most common)\n\n```bash\naws ec2 describe-security-groups --group-ids \u003Crds-sg-id> --output json | \\\n  jq('.SecurityGroups[0].IpPermissions')\n```\n\nThe RDS SG must have an inbound rule allowing the client's SG (or\nCIDR) on the DB port (5432 for Postgres, 3306 for MySQL, 1433 for\nSQL Server, 1521 for Oracle, 5439 for Redshift). If the rule is\nmissing or scoped to a different SG \u002F CIDR, **ROOT_CAUSE_IDENTIFIED**\nwith `ROOT_CAUSE: NETWORK_SG_INBOUND`.\n\n#### 1b: Client SG egress (the other half of the bidirectional check)\n\n```bash\naws ec2 describe-security-groups --group-ids \u003Cclient-sg-id> --output json | \\\n  jq('.SecurityGroups[0].IpPermissionsEgress')\n```\n\nThe client SG must allow egress to the RDS endpoint on the DB port.\nThe default egress is allow-all (0.0.0.0\u002F0); a locked-down SG with a\nrestricted egress (e.g., only 443) silently drops the DB-bound SYN. If\nthe egress rule is missing, **ROOT_CAUSE_IDENTIFIED** with\n`ROOT_CAUSE: NETWORK_SG_EGRESS`.\n\n#### 1c: NACL and route table (deeper network)\n\nIf both SGs allow the flow, check the NACL on both subnets (NACLs are\nstateless — both inbound and outbound rules must allow the flow in\nboth directions) and the route table (for cross-VPC peering \u002F TGW\nroutes). Fold to `NETWORK_NACL` or `NETWORK_ROUTE`.\n\n### Step 2: Connection refused — instance-level failure\n\nSymptom: `connection refused` (RST returned, not a timeout). The\nnetwork path is open but the DB engine is not accepting connections.\n\n| describe-db-instances signal | ROOT_CAUSE |\n|---|---|\n| `DBInstanceStatus: available` but `StorageStatus: storage-full` | `INSTANCE_STORAGE_FULL` — enable storage auto-scaling |\n| `DBInstanceStatus: modifying` | `INSTANCE_MAINTENANCE` — wait for the window to complete |\n| Recent event `DB instance restarted` \u002F `Multi-AZ failover completed` | `INSTANCE_FAILOVER` — the writer changed; application DNS cache may be stale |\n| `DBInstanceStatus: stopped` \u002F `starting` | `INSTANCE_UNAVAILABLE` — start the instance |\n| `DBInstanceStatus: incompatible-*` | Fold to Step 9 (parameter or option group) |\n\nFor `storage-full`: confirm with `describe-db-instances` showing\n`StorageStatus: storage-full`. The fix is:\n\n```bash\naws rds modify-db-instance --db-instance-identifier \u003Cid> \\\n  --storage-auto-scaling --max-allocated-storage \u003CN> --apply-immediately\n```\n\n### Step 3: Authentication failure — credentials vs IAM DB auth\n\nSymptom: `FATAL: password authentication failed for user \"...\"`,\n`Access denied for user`, or `28000: authentication failed`.\n\n#### 3a: Master \u002F static credentials\n\nIf the connection uses a static password (Secrets Manager, hardcoded,\nor env var), the credentials are wrong, rotated, or stale. Check\nSecrets Manager:\n\n```bash\naws secretsmanager get-secret-value --secret-id \u003Csecret-id> \\\n  --output json | jq('.SecretString')\n```\n\nCompare the username \u002F password against what the application is using.\nIf the secret was rotated and the application cached the old value,\n**ROOT_CAUSE_IDENTIFIED** with `ROOT_CAUSE: AUTH_CREDENTIALS`.\n\n#### 3b: IAM database auth\n\nIf `IAMDatabaseAuthenticationEnabled: true` and the application uses\nan SigV4 token as the password, the token expires after 15 minutes.\nCheck:\n\n1. The IAM policy attached to the application's role must include\n   `rds-db:connect` on the DB instance ARN.\n2. The database user must exist with\n   `AWSAuthenticationPlugin: rds_iam` (Postgres) or\n   `AWSAuthenticationPlugin: mysql_native_password` (MySQL).\n3. The token must be regenerated before each connection (or at least\n   every 14 minutes).\n\n```bash\n# Verify the IAM policy grants rds-db:connect\naws iam simulate-principal-policy \\\n  --policy-source-arn \u003Capp-role-arn> \\\n  --action-names rds-db:connect \\\n  --resource-arns arn:aws:rds-db:\u003Cregion>:\u003Caccount>:dbuser:\u003Cinstance>\u002F\u003Cuser> \\\n  --output json\n```\n\nIf `implicitDeny`, **ROOT_CAUSE_IDENTIFIED** with\n`ROOT_CAUSE: AUTH_IAM_DB`. If the token is older than 15 minutes,\n`ROOT_CAUSE: AUTH_EXPIRED`.\n\n### Step 4: SSL \u002F TLS errors\n\nSymptom: `SSL connection required`, `TLS handshake failed`,\n`no SSL capabilities`, or `certificate verify failed`.\n\n#### 4a: force_ssl \u002F require_secure_transport\n\n```bash\naws rds describe-db-instances --db-instance-identifier \u003Cid> \\\n  --output json | jq('.DBInstances[0] | .DBParameterGroups')\n```\n\nIf the parameter group has `rds.force_ssl: 1` (Postgres) or\n`require_secure_transport: 1` (MySQL) and the client connects without\n`sslmode=require` (or the MySQL equivalent), the connection is\nrejected. **ROOT_CAUSE_IDENTIFIED** with `ROOT_CAUSE: TLS_CONFIG`.\n\n#### 4b: CA bundle mismatch\n\n```bash\naws rds describe-db-instances --db-instance-identifier \u003Cid> \\\n  --output json | jq('.DBInstances[0].CACertificateIdentifier')\n```\n\nIf the instance uses `rds-ca-e2022` (the recommended rotation) but the\nclient's trust store has `rds-ca-2019`, the TLS handshake fails with\n`certificate verify failed`. **ROOT_CAUSE_IDENTIFIED** with\n`ROOT_CAUSE: TLS_CERT`. Fix: update the client trust store with the\nnew CA bundle from the AWS RDS docs.\n\n### Step 5: Too many connections — capacity\n\nSymptom: `FATAL: too many connections`,\n`FATAL: remaining connection slots are reserved for non-replication\nsuperuser connections`, or connection pool exhaustion at the client.\n\n#### 5a: max_connections (RDS for MySQL \u002F Postgres)\n\n```bash\naws rds describe-db-instances --db-instance-identifier \u003Cid> \\\n  --output json | jq('.DBInstances[0].DBParameterGroups')\n```\n\nIf a parameter group overrides `max_connections` below the default,\n**ROOT_CAUSE_IDENTIFIED** with\n`ROOT_CAUSE: CAPACITY_MAX_CONNECTIONS` (or\n`PARAM_GROUP_OVERRIDE` if the override is the root issue).\n\n#### 5b: Aurora derived limit\n\nAurora derives `max_connections` from the instance class\n(`LEAST({DBInstanceClassMemory\u002F9531392, 5000})` for Aurora\nPostgreSQL). Upsizing the instance class raises the limit\nautomatically. If the application exceeds the derived limit,\n**ROOT_CAUSE_IDENTIFIED** with `ROOT_CAUSE: CAPACITY_DERIVED_LIMIT`.\nFix: upsize the instance class OR add an RDS Proxy to pool\nconnections.\n\n#### 5c: RDS Proxy not absorbing the load\n\nIf an RDS Proxy is configured but the application still hits\n`too many connections`, the proxy may be misconfigured (wrong target\nSG, wrong secrets, or the application is bypassing the proxy\nendpoint). Fold to Step 8.\n\n### Step 6: DNS resolution — custom vs cluster endpoint\n\nSymptom: `could not translate host name`, `nodename nor servname`,\nor the application connects to the wrong instance after a failover.\n\n#### 6a: Custom endpoint vs cluster endpoint\n\nAurora custom endpoints route to a subset of instances; the cluster\nendpoint follows the writer; the reader endpoint round-robins across\nreaders. If the application uses a custom endpoint for writes but the\ncustom endpoint only includes readers, every write fails with\n`cannot execute INSERT in a read-only transaction`.\n\n```bash\naws rds describe-db-clusters --db-cluster-identifier \u003Ccluster> \\\n  --output json | jq('.DBClusters[0] | {\n    CustomEndpoints: .CustomEndpoints,\n    Endpoint, ReaderEndpoint,\n    Members: [.DBClusterMembers[] | {DBInstanceIdentifier,\n      IsClusterWriter}]}')\n```\n\nConfirm the endpoint includes the writer. **ROOT_CAUSE_IDENTIFIED**\nwith `ROOT_CAUSE: DNS_CUSTOM_ENDPOINT` or `DNS_CLUSTER_ENDPOINT`.\n\n#### 6b: Stale DNS after failover\n\nStale-DNS-after-failover guidance moved verbatim to\n[references\u002Fconnectivity-layer-reference.md](references\u002Fconnectivity-layer-reference.md).\n\n### Step 7: Replication lag — read replica and global database\n\n#### 7a: Read replica lag\n\n```bash\naws cloudwatch get-metric-statistics --namespace AWS\u002FRDS \\\n  --metric-name ReplicaLag \\\n  --dimensions Name=DBInstanceIdentifier,Value=\u003Creplica-id> \\\n  --start-time $(date -d '-1 hour' +%FT%TZ) --end-time $(date +%FT%TZ) \\\n  --period 300 --statistics Average,Maximum --output json\n```\n\n`ReplicaLag` > 30s and rising indicates the replica cannot keep up\nwith the source. Reads return stale data. **ROOT_CAUSE_IDENTIFIED**\nwith `ROOT_CAUSE: REPLICATION_LAG`. Fix: upsize the replica, reduce\nwrite load on the source, or route reads elsewhere.\n\n#### 7b: Aurora global database replication\n\nFor Aurora global databases, the secondary cluster replicates from\nthe primary over a dedicated channel. If the channel is interrupted\n(network, AWS-side issue) or the secondary cannot keep up,\n`ReplicationLag` on the secondary climbs. Check the secondary cluster\nstatus and the AWS Health dashboard for cross-region issues.\n**ROOT_CAUSE_IDENTIFIED** with `ROOT_CAUSE: GLOBAL_DB_REPLICATION`.\n\n### Step 8: RDS Proxy connectivity\n\nRDS Proxy probe commands and the common proxy failure pattern table moved verbatim to\n[references\u002Fengine-auth-reference.md](references\u002Fengine-auth-reference.md).\n\n### Step 9: Parameter group and option group misconfiguration\n\n#### 9a: Parameter group override\n\nThe parameter-override culprit table moved verbatim to\n[references\u002Fengine-auth-reference.md](references\u002Fengine-auth-reference.md).\n\n#### 9b: Option group conflict\n\nOption-group conflict patterns moved verbatim to\n[references\u002Fengine-auth-reference.md](references\u002Fengine-auth-reference.md).\n\n### Step 10: INSUFFICIENT_DATA — when to bail\n\nEmit `INSUFFICIENT_DATA` when:\n\n- The input lacks the instance identifier or endpoint and the symptom\n  is too generic.\n- A probe requires operator input (e.g., the application's connection\n  string is unknown).\n- The symptom matches no row in the quick navigation table and the\n  describe-db-instances output is empty or stale.\n- An AWS Health event is OPEN and may be the cause.\n\nALWAYS list the missing inputs and the exact next probe to run.\n\n## Output format\n\nSee the STRICT output contract above. Two worked examples follow.\n\n### Worked example — connection timeout, RDS SG missing inbound\n\n```text\nTARGET: db-prod-postgres (endpoint: db-prod-postgres.abc123.us-east-1.rds.amazonaws.com:5432)\nVERDICT: ROOT_CAUSE_IDENTIFIED\nROOT_CAUSE: NETWORK_SG_INBOUND\nREASON: The RDS SG (sg-rds-aaa) has no inbound rule allowing the\n  application's SG (sg-app-bbb) on port 5432. The only inbound rule\n  permits sg-old-ccc (a decommissioned SG). The client SG egress is\n  allow-all (0.0.0.0\u002F0), so the packet reaches the RDS subnet but is\n  dropped at the RDS SG.\nEVIDENCE:\n  - Probe: aws ec2 describe-security-groups --group-ids sg-rds-aaa\n    returns IpPermissions with one rule: port 5432 from sg-old-ccc —\n    sg-app-bbb is absent.\n  - Probe: aws ec2 describe-security-groups --group-ids sg-app-bbb\n    returns IpPermissionsEgress: 0.0.0.0\u002F0 all ports (egress is open).\n  - Passing: describe-db-instances shows DBInstanceStatus: available,\n    StorageStatus: healthy (instance is not the issue); the client is\n    in the same VPC (route table is not the issue).\nREMEDIATION:\n  1. Add an inbound rule to sg-rds-aaa allowing sg-app-bbb on 5432:\n     aws ec2 authorize-security-group-ingress --group-id sg-rds-aaa \\\n       --protocol tcp --port 5432 --source-security-group-id sg-app-bbb\n  2. Verify from the client host: psql -h \u003Cendpoint> -U \u003Cuser> -d \u003Cdb>\n     (should connect within seconds).\nCONFIRM: Before adding the rule, emit and await:\n  \"CONFIRM: About to add inbound 5432 from sg-app-bbb to sg-rds-aaa.\n   Proceed? (yes\u002Fno)\"\n```\n\n### Worked example — INSUFFICIENT_DATA\n\nThe INSUFFICIENT_DATA worked example moved verbatim to\n[references\u002Fworked-examples.md](references\u002Fworked-examples.md).\n\n## Pre-flight safety checks\n\nPre-flight safety checks (confirm gate, read-only-first, disruption notes,\nbatch limit) moved verbatim to [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md).\n\n## Remediation guidance\n\n| ROOT_CAUSE | Specific fix |\n|---|---|\n| `NETWORK_SG_INBOUND` | Add inbound rule to the RDS SG for the client SG \u002F CIDR on the DB port. |\n| `NETWORK_SG_EGRESS` | Add egress rule to the client SG for the RDS endpoint on the DB port. |\n| `NETWORK_NACL` \u002F `NETWORK_ROUTE` | Add NACL \u002F route table rule for bidirectional flow. |\n| `NETWORK_SUBNET_GROUP` | Add a subnet in the client's AZ to the DB subnet group. |\n| `INSTANCE_STORAGE_FULL` | Enable storage auto-scaling: `modify-db-instance --storage-auto-scaling --max-allocated-storage \u003CN>`. |\n| `INSTANCE_FAILOVER` | Wait for the cluster endpoint DNS to update; flush the app DNS cache. |\n| `AUTH_CREDENTIALS` | Rotate \u002F update the secret in Secrets Manager; refresh the app's cached value. |\n| `AUTH_IAM_DB` | Add `rds-db:connect` to the app role; create the DB user with `AWSAuthenticationPlugin`. |\n| `TLS_CONFIG` | Add `sslmode=require` (Postgres) or `requireSSL=true` (MySQL) to the connection string. |\n| `TLS_CERT` | Update the client trust store with the CA bundle matching `CACertificateIdentifier`. |\n| `CAPACITY_MAX_CONNECTIONS` \u002F `CAPACITY_DERIVED_LIMIT` | Upsize the instance class OR add an RDS Proxy to pool connections. |\n| `CAPACITY_PROXY` | Fix the proxy SG, secrets, or application connection string. |\n| `REPLICATION_LAG` | Upsize the replica, reduce source write load, or route reads elsewhere. |\n| `GLOBAL_DB_REPLICATION` | Check the secondary cluster status and AWS Health; recover or re-establish the channel. |\n| `DNS_CUSTOM_ENDPOINT` \u002F `DNS_CLUSTER_ENDPOINT` | Add the writer to the custom endpoint; flush DNS; use the writer endpoint for writes. |\n| `PARAM_GROUP_OVERRIDE` | Revert or correct the parameter; reboot if required. |\n| `OPTION_GROUP_CONFLICT` | Resolve the option conflict; re-apply the option group. |\n\n## References (load on demand)\n\n- [references\u002Fadvanced-patterns.md](references\u002Fadvanced-patterns.md) — expert-heuristic deep dives (bidirectional SG check order, Aurora writer\u002Freader routing, storage-full) and the configuration dependency graph (moved from this file)\n- [references\u002Fdiagnostic-commands.md](references\u002Fdiagnostic-commands.md) — Step 0 pre-flight gathering commands (probes 1-4) and the pre-flight safety checks (moved from this file)\n- [references\u002Fworked-examples.md](references\u002Fworked-examples.md) — the INSUFFICIENT_DATA worked example; the primary connection-timeout example stays in this file\n- [references\u002Fengine-auth-reference.md](references\u002Fengine-auth-reference.md) — IAM DB auth, TLS\u002FCA rotation, parameter and option group, storage-full, proxy, DNS detail; extended with Step 8 proxy probes, the 9a parameter-override table, and 9b option-group conflicts\n- [references\u002Fconnectivity-layer-reference.md](references\u002Fconnectivity-layer-reference.md) — engine port registry, SG\u002FNACL evaluation rules, Aurora endpoint behaviour, cross-VPC matrix, AWS Health event categories; extended with the 6b stale-DNS-after-failover guidance\n\n## Domain\n\nAWS CloudOps \u002F RDS & Aurora Databases, Connectivity Diagnostics, VPC\nNetworking for Databases, IAM Database Authentication, TLS \u002F SSL,\nStorage Auto-Scaling, and Read Replica \u002F Global Database Replication.\n\n## AWS documentation\n\n- RDS troubleshooting — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FCHAP_Troubleshooting.html\n- Security group rules — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_VPC.WorkingWithRDSInstanceinaVPC.html\n- IAM database auth — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUsingWithRDS.IAMDBAuth.html\n- SSL\u002FTLS — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUsingWithRDS.SSL.html\n- Storage auto-scaling — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002FUSER_PIOPS.StorageTypes.html#USER_PIOPS.Autoscaling\n- Aurora endpoints — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002FAurora.Overview.Endpoints.html\n- Aurora global databases — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002Faurora-global-database.html\n- RDS Proxy — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FUserGuide\u002Frds-proxy.html\n- Aurora failover — https:\u002F\u002Fdocs.aws.amazon.com\u002FAmazonRDS\u002Flatest\u002FAuroraUserGuide\u002FConcepts.AuroraHighAvailability.html\n",{"name":5,"license":15,"metadata":24,"description":7,"compatibility":43},{"tags":25,"phase":26,"author":27,"domain":28,"family":29,"version":30,"keywords":31,"task_type":32,"complexity":33,"entry_point":34,"skill_class":35,"when_to_use":36,"requires_llm":37,"verdict_shape":38,"when_not_to_use":39,"lifecycle_status":40,"invocation_schema":41,"supports_pipeline":37,"activation_triggers":42},"rds, aurora, databases, troubleshoot, connectivity, security-group, tls, iam-db-auth, storage-full","2","Jacky Chan — AWS Community Builder","aws-cloudops","Databases","0.1.0","RDS, Aurora, connection timeout, security group, bidirectional SG, subnet group, DB subnet group, DNS resolution, custom endpoint, cluster endpoint, writer endpoint, reader endpoint, IAM database auth, SSL, TLS, force_ssl, require_secure_transport, CA bundle, rds-ca-2019, rds-ca-e2022, max_connections, Aurora derived limit, storage-full, storage auto-scaling, CPU exhaustion, memory exhaustion, read replica lag, Aurora global database, RDS Proxy, parameter group, option group","troubleshoot","high","false","capability","Diagnosing why an application cannot reach an RDS or Aurora database — connection timeout (SG, subnet, NACL, route table), authentication failure (master credentials, IAM database auth, expired password), SSL\u002FTLS handshake error (force_ssl, CA bundle, TLS version), connection refused (instance unavailable, Multi-AZ failover, maintenance window, storage-full), too many connections (max_connections, Aurora derived limit, RDS Proxy), high latency (CPU\u002Fmemory exhaustion, read replica lag), DNS resolution failure (custom endpoint, cluster endpoint, cross-region), or RDS Proxy connectivity issues.","true","ROOT_CAUSE_IDENTIFIED | INSUFFICIENT_DATA","Query-level performance tuning (use Performance Insights + slow query log); RDS instance configuration posture audits (use rds-instance-auditor); Aurora failover automation (use aurora-failover-operator); RDS backup and restore operations (use rds-backup-restore-operator); parameter group authoring (use rds-parameter-group-deployer).","active","Input: either (a) a symptom description (the error string or observed behaviour, the DB instance or cluster identifier), OR (b) a live-account scenario where the agent runs aws rds describe-db-instances \u002F describe-db-clusters \u002F describe-events \u002F describe-security-groups to gather evidence. Output: a deterministic TARGET \u002F VERDICT \u002F ROOT_CAUSE \u002F REASON \u002F EVIDENCE \u002F REMEDIATION block where VERDICT ∈ {ROOT_CAUSE_IDENTIFIED, INSUFFICIENT_DATA} and ROOT_CAUSE names the specific failure category (NETWORK_SG_INBOUND \u002F NETWORK_SG_EGRESS \u002F NETWORK_NACL \u002F NETWORK_ROUTE \u002F NETWORK_SUBNET_GROUP \u002F NETWORK_CROSS_VPC \u002F DNS_CUSTOM_ENDPOINT \u002F DNS_CLUSTER_ENDPOINT \u002F AUTH_CREDENTIALS \u002F AUTH_IAM_DB \u002F AUTH_EXPIRED \u002F TLS_CONFIG \u002F TLS_CERT \u002F INSTANCE_UNAVAILABLE \u002F INSTANCE_FAILOVER \u002F INSTANCE_MAINTENANCE \u002F INSTANCE_STORAGE_FULL \u002F CAPACITY_MAX_CONNECTIONS \u002F CAPACITY_DERIVED_LIMIT \u002F CAPACITY_PROXY \u002F RESOURCE_CPU \u002F RESOURCE_MEMORY \u002F REPLICATION_LAG \u002F GLOBAL_DB_REPLICATION \u002F PARAM_GROUP_OVERRIDE \u002F OPTION_GROUP_CONFLICT \u002F UNKNOWN).","RDS connection timeout, cannot connect to RDS, Aurora connection failed, authentication failed for database, FATAL password authentication failed, Access denied for user, SSL connection required, TLS handshake failed RDS, connection refused RDS, too many connections RDS, FATAL too many connections, RDS storage-full, RDS CPU exhaustion, read replica lag, Aurora global database replication, RDS Proxy unreachable, parameter group max_connections override, option group conflict, RDS DNS not resolving","Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). Offline symptom classification works from pasted error messages and instance metadata. Live-account diagnosis uses aws rds describe-db-instances, describe-db-clusters, describe-events, describe-db-subnet-groups, describe-db-parameter-groups, describe-option-groups, aws ec2 describe-security-groups, describe-network-acls, describe-route-tables, describe-vpc-endpoints, aws logs get-log-events \u002F filter-log-events...","main","eeace51bd5dd1db2030729d83eeab988861553aa","skills\u002Frds-connectivity-troubleshooter",0,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills","https:\u002F\u002Fgithub.com\u002Fjhkchan","CTO & Co-Founder @Beever-AI  & @votee. CAISP. AWS Community Builder. Organizer@GDGCloudHK. MongoDB Champion. Technopreneur + Infornographer.","2026-09-24 08:54:10",{"source":8,"skillName":5,"commands":53,"prompt":63,"zipAvailable":64},[54,57,60],{"tool":55,"command":56},"npx","npx skills add jhkchan\u002Faws-agent-skills --skill rds-connectivity-troubleshooter",{"tool":58,"command":59},"bunx","bunx skills add jhkchan\u002Faws-agent-skills --skill rds-connectivity-troubleshooter",{"tool":61,"command":62},"pnpm","pnpm dlx skills add jhkchan\u002Faws-agent-skills --skill rds-connectivity-troubleshooter","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002FSKILL.md",true,[66,72,76,80,84,88,92,96,100,104,108,112,116,120,124,128,132,136,140,144],{"path":67,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":70,"sourceHtmlUrl":71},"eval\u002Ftest-cases.yaml","REFERENCE","参考文档",8362,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Feval\u002Ftest-cases.yaml",{"path":73,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":74,"sourceHtmlUrl":75},"evals\u002Fbaselines\u002Fauth-failed-iam-token-expiry.md",579,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fbaselines\u002Fauth-failed-iam-token-expiry.md",{"path":77,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":78,"sourceHtmlUrl":79},"evals\u002Fbaselines\u002Fconnection-refused-failover.md",420,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fbaselines\u002Fconnection-refused-failover.md",{"path":81,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":82,"sourceHtmlUrl":83},"evals\u002Fbaselines\u002Ftimeout-sg-missing-rule.md",736,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fbaselines\u002Ftimeout-sg-missing-rule.md",{"path":85,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":86,"sourceHtmlUrl":87},"evals\u002Fbaselines\u002Ftls-cert-mismatch.md",416,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fbaselines\u002Ftls-cert-mismatch.md",{"path":89,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":90,"sourceHtmlUrl":91},"evals\u002Fbaselines\u002Ftoo-many-connections-aurora-derived.md",491,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fbaselines\u002Ftoo-many-connections-aurora-derived.md",{"path":93,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":94,"sourceHtmlUrl":95},"evals\u002Feval_config.yaml",2016,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Feval_config.yaml",{"path":97,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":98,"sourceHtmlUrl":99},"evals\u002Fevals.json",4864,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fevals.json",{"path":101,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":102,"sourceHtmlUrl":103},"evals\u002Fprompts\u002Fauth-failed-iam-token-expiry.md",1261,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fprompts\u002Fauth-failed-iam-token-expiry.md",{"path":105,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":106,"sourceHtmlUrl":107},"evals\u002Fprompts\u002Fconnection-refused-failover.md",946,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fprompts\u002Fconnection-refused-failover.md",{"path":109,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":110,"sourceHtmlUrl":111},"evals\u002Fprompts\u002Ftimeout-sg-missing-rule.md",1266,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fprompts\u002Ftimeout-sg-missing-rule.md",{"path":113,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":114,"sourceHtmlUrl":115},"evals\u002Fprompts\u002Ftls-cert-mismatch.md",1181,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fprompts\u002Ftls-cert-mismatch.md",{"path":117,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":118,"sourceHtmlUrl":119},"evals\u002Fprompts\u002Ftoo-many-connections-aurora-derived.md",955,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fevals\u002Fprompts\u002Ftoo-many-connections-aurora-derived.md",{"path":121,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":122,"sourceHtmlUrl":123},"examples\u002FREADME.md",5850,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fexamples\u002FREADME.md",{"path":125,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":126,"sourceHtmlUrl":127},"references\u002Fadvanced-patterns.md",3845,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Freferences\u002Fadvanced-patterns.md",{"path":129,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":130,"sourceHtmlUrl":131},"references\u002Fconnectivity-layer-reference.md",8973,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Freferences\u002Fconnectivity-layer-reference.md",{"path":133,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":134,"sourceHtmlUrl":135},"references\u002Fdiagnostic-commands.md",2570,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Freferences\u002Fdiagnostic-commands.md",{"path":137,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":138,"sourceHtmlUrl":139},"references\u002Fengine-auth-reference.md",12796,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Freferences\u002Fengine-auth-reference.md",{"path":141,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":142,"sourceHtmlUrl":143},"references\u002Fworked-examples.md",862,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Freferences\u002Fworked-examples.md",{"path":145,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":146,"sourceHtmlUrl":147},"skill.usf.yaml",4633,"https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Frds-connectivity-troubleshooter\u002Fskill.usf.yaml",[149,156,163,170,177,184],{"skillNo":150,"name":151,"nameZh":6,"description":152,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":153,"categories":154,"keywords":155,"publishedAt":19},"01M7ADVD60X4PCAPKSE58P9NDC","xray-tracing-deployer","Deploys AWS X-Ray distributed tracing with production-grade configuration: X-Ray daemon deployment (EC2 systemd, ECS sidecar, EKS DaemonSet, Lambda built-in), SDK instrumentation (Java, Python, Node.js, Go, .NET) with annotations and metadata, sampling rules (default reservoir + rate vs custom rules with URL \u002F method \u002F service-name predicates), service map generation, trace query and retrieval, groups and insights, CloudWatch ServiceLens integration, IAM permissions (PutTraceSegments, GetSamplingRules), and latest features (Lambda Powertools tracing, OpenTelemetry \u002F ADOT Collector support, W3C trace context). Emits a READY_TO_DEPLOY checklist with every configuration item verified. Use when enabling X-Ray on a new application, adding tracing to ECS \u002F EKS \u002F Lambda, validating a sampling configuration, or generating daemon deployment CLI and IaC templates. Triggers: X-Ray, distributed tracing, X-Ray daemon, sampling rules, service map, X-Ray SDK, OpenTelemetry, ADOT, Lambda Powertools, CloudWatch ServiceLens.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fxray-tracing-deployer\u002FSKILL.md",[],[],{"skillNo":157,"name":158,"nameZh":6,"description":159,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":160,"categories":161,"keywords":162,"publishedAt":19},"01M7ADTVJ7DCBM9CB1FY51GH7J","wellarchitected-workload-auditor","Audits AWS Well-Architected Tool workloads for review staleness, per-pillar high-risk issue counts, milestone tracking gaps, and remediation plan completeness. Emits a deterministic verdict (STALE_REVIEW | HIGH_RISK | CONFIG_GAP | OK) per workload with per-pillar risk breakdown, milestone trend analysis, and specific remediation. Use when reviewing Well-Architected workloads, checking WA review freshness, auditing high-risk issues per pillar, validating milestone coverage, or tracking remediation plans.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwellarchitected-workload-auditor\u002FSKILL.md",[],[],{"skillNo":164,"name":165,"nameZh":6,"description":166,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":167,"categories":168,"keywords":169,"publishedAt":19},"01M7ADT59G6FTTNZYHT2QRPNNP","wellarchitected-review-operator","Operates AWS Well-Architected Tool reviews end-to-end — creates workloads, runs pillar reviews (operational excellence, security, reliability, performance, cost optimization, sustainability, Prosperity), answers questions with risk-tier rationale, generates reports and improvement plans, creates milestones, and integrates with Trusted Advisor and Well-Architected Labs. Runs deterministic pre-checks (workload existence, lens availability, pillar coverage), emits the exact update-answer \u002F get-consolidated-report CLI behind a CONFIRM gate, and verifies risk-tier transitions. Emits READY | BLOCKED | COMPLETED. Use when creating a workload, answering pillar questions, generating an improvement plan, creating a milestone, or integrating with Trusted Advisor. Triggers: well-architected tool, well architected review, workload review, pillar, operational excellence, security, reliability, performance, cost optimization, sustainability, prosperity, improvement plan, milestone, trusted advisor.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwellarchitected-review-operator\u002FSKILL.md",[],[],{"skillNo":171,"name":172,"nameZh":6,"description":173,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":174,"categories":175,"keywords":176,"publishedAt":19},"01M7ADSH5JA9QHNPPHQCD7EVA7","wafv2-web-acl-deployer","Provisions AWS WAFv2 Web ACLs with correct production defaults: scope selection (CLOUDFRONT vs REGIONAL), managed rule groups (AWSManagedRulesCommonRuleSet, SQLiRuleSet, LinuxRuleSet, WindowsRuleSet, IPReputationList, BotControlRuleSet, AmazonIpReputationList, ATP), custom rules (byte match, IP set, regex pattern, geo match, size constraint, rate-based), rule priority ordering, logging destinations (Kinesis Firehose, CloudWatch Logs, S3), resource association (ALB, API Gateway, CloudFront), rate-based rules with aggregate keys (IP, forwarded IP), Challenge and CAPTCHA actions, and account takeover prevention. Emits a READY_TO_DEPLOY checklist. Use when creating a Web ACL, attaching managed rules, configuring rate limiting, enabling WAF logging, associating a Web ACL with an ALB\u002FAPI Gateway\u002FCloudFront, or configuring CAPTCHA\u002FChallenge actions. Triggers: create WAF Web ACL, WAFv2, managed rule groups, WAF rate limiting, WAF logging, WAF CAPTCHA, WAF Challenge action, bot control, ATP managed rules.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwafv2-web-acl-deployer\u002FSKILL.md",[],[],{"skillNo":178,"name":179,"nameZh":6,"description":180,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":181,"categories":182,"keywords":183,"publishedAt":19},"01M7ADRYFAT8R73FBNCHRC573X","wafv2-web-acl-auditor","Audits AWS WAFv2 Web ACL configurations to determine whether the ACL provides effective protection: default-action posture (Allow vs Block), managed-rule-group coverage gaps, rule effectiveness (BLOCK vs COUNT, shadow\u002Fbypass rules, stale exclusions), rate-based rule correctness, logging and visibility configuration, and text-transformation bypass vectors. Emits a deterministic verdict per Web ACL. Use when reviewing a WAFv2 Web ACL for security posture, checking managed-rule coverage, validating rate-limit configuration, auditing WAF logging, or hardening a WAF before production deployment. Triggers: WAF, WAFv2, Web ACL, managed rule group, AWSManagedRules, rate-based rule, CAPTCHA, Challenge, OverrideAction, Count mode, forwarded IP, Bot Control, ATP, text transformation, WAF logging, WAF visibility, OWASP, SQLi, XSS, RFI.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwafv2-web-acl-auditor\u002FSKILL.md",[],[],{"skillNo":185,"name":186,"nameZh":6,"description":187,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":188,"categories":189,"keywords":190,"publishedAt":19},"01M7ADR8W7G72JX4KFGN03QK8Q","waf-rule-deployer","Provisions AWS WAFv2 rule sets with production defaults: Web ACL creation (CloudFront scope in us-east-1 vs Regional scope), rule groups (managed vs custom) with capacity planning in WCU (Web ACL Capacity Units, 1500 default per ACL), managed rule groups (AWSManagedRulesCommonRuleSet, AWSManagedRulesSQLiRuleSet, AWSManagedRulesAmazonIpReputationList, AWSManagedRulesLinuxRuleSet, AWSManagedRulesWindowsRuleSet, AWSManagedRulesUnixRuleSet), custom rules (byte-match, regex pattern set, geo-match, rate-based), IP set and regex pattern set resources, rule priority ordering (lowest number evaluated first), action types (allow, block, count, CAPTCHA, Challenge), bot control managed rule group, account takeover. Triggers: create WAF rule, WAF managed rule group, WAF custom rule, WAF byte match, WAF geo match, WAF rate based rule, WAF IP set, WAF regex pattern set, WAF WCU budget, WAF CloudFront ACL, WAF logging Kinesis Firehose, WAF bot control, WAF ATP, WAF label matching, WAF Challenge action, WAF CAPTCHA.","https:\u002F\u002Fgithub.com\u002Fjhkchan\u002Faws-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwaf-rule-deployer\u002FSKILL.md",[],[]]